Abstract
To date, identification performance from composites remains poor, especially where forensically valid procedures are adopted in the construction process. Several experiments have assessed some of the reasons for poor performance. In the first experiment, the effects of operator performance were assessed through the construction of composites of the same targets across novice and experienced operators. Performance was assessed through uncued naming and likeness ratings and results indicated that the performance of the experienced operator may have been no better than novice operator performance (however, there were procedural differences). Target distinctiveness was also manipulated and targets rated as being highly distinctive were identified more often than targets with low distinctiveness. The second experiment concentrated on the effects of retention interval (2 days and 1 week) and artistic enhancement. Naming and likeness ratings were poor. Likeness ratings revealed an advantage for composites constructed with the shorter retention interval. The use of artistic elaboration appeared advantageous with the longer retention interval of one week. A third experiment implemented retention intervals of 3 – 4 hours and 2 days. Again, naming levels and likeness ratings were poor. There was a trend in the direction of the shorter retention interval providing better identification results.